WRF stands for the Weather and Research Forecast Model. Pretty much, it's thousands of lines of FORTRAN code that, when executed, predict the weather. This is known as numerical weather prediction (NWP). Learning a little about NWP will help you appreciate your cute little weather app on your phone.
I'm in Boulder at the University Corporation for Atmospheric Research (UCAR) for the WRF Tutorial Training. There are at least 40 people here from all kinds of professions and literally from every continent (yes, even Antarctica is represented. A researcher from England is headed to the South Pole later this year.)
To run the WRF model, all you need is a Linux computer. You can download the model for free and then make your own weather forecasts for any area in the world you are interested in. Ok, it's a little more difficult than that. That's why I'm here for a week.
A weather forecast first starts as data. You need weather observations taken from weather stations, weather balloons, and satellites. Oh, and you also need another weather forecast. That piece might seem unnecessary. Why do you need a forecast to make a forecast?
The GFS (global) and NAM (North America) model data is created at low resolution. Have you ever seen a pixelated picture? That's what the GFS model looks like.
When you increase the resolution, the picture looks much more realistic...
Obviously, the second picture of Mario is better. That is why WRF is used. Weather forecasts made by WRF use the coarse forecasts to increase the resolution of the forecast. These forecasts can be tuned for specific purposes.
WRF looks good, so why don't we just use that? Or why not run the GFS or NAM at higher resolutions? Well, there are several problems. The most obvious problem is computing power. Computers aren't powerful or fast enough to create a global weather forecast at this high resolution.
WRF is like a weather playground. The code allows the user to change anything you want. Imagine, all the knobs and buttons that control the weather in your control!!! Have you ever wondered what the weather would be like if there were no mountains? What if the lake was 5 degrees warmer? How can we make it snow? How does the weather change after a volcano erupts? Why was the forecast wrong and how can we get the forecast right next time?
Why is this useful? Scientists can investigate ways to make weather forecasts more accurate during specific situations. There is a high resolution WRF model run over Utah, and you can get the data from the National Weather Service. You can get the WRF forecast for Utah here: http://www.wrh.noaa.gov/slc/projects/wrf/wrf.html. The trick is interpreting all this data. That's why meteorologist have jobs.
For my research the the University of Utah, I'll be using WRF to better understand the dreaded inversions and cold air pools that cause unhealthy air quality in the winters. Inversions are not accurately modeled in large scale models like the GFS or NAM. Hopefully with the WRF we'll learn how to better forecast these events.
And this is Warf, who joined us for our WRF lectures...